Basic AWS EC2 SQL High Availability Query

Question

We are considering to move our application and SQL servers to AWS EC2 instances. Currently, we have one SQL Server Standard and we have failover clustering for high availability.

On AWS, We are planning to have EC2 instance(s) with EBS volumes as needed for SQL server. So, to sql server implement high availability only option available if Synchronous DB Mirroring unless we go for some storage level replication and implement cluster on top of it.

This application does not need cross availability zone (multi site) high availability.

My basic question is why we need to take care of SQL High availability? Asking this question as

1. Keeping EC2 up is the amazon's responsibility
2. EBS volumes gets automatically replicated internally in availability zone.

Only think I can think of that we need separate high availability is EC2 instance becomes irresponsive or something on OS / Driver side gets corrupted. Do you see any other reason then these ones?

Thanks for taking time to read this question.

Answer 1

If you'd like a managed, highly available SQL Server, Amazon RDS would be a better option, as you would have HA (Multi-AZ), Backup, Patching and other administrative tasks automatically. Using RDS, you do not have to manage the underlying EC2 instances and EBS volumes yourself.

Using Amazon EC2, you need to take care of application HA (SQL Server in this case) because this is not (and can not be) managed by Amazon. For example, if you are hosting your personal blog you could use one instance in a single Availability Zone, for a database with primary-secondary (HA) replication you may use two instances in separate AZs and for a multi-master distributed database, you may use several instances in several regions. The design and cost of High Availability is very context dependent.

You can try and learn RDS for free with this self-paced lab: https://run.qwiklab.com/focuses/preview/1013?locale=en